Reproduce “HOForce: Motion and Force Reconstruction of Hand-Object Interaction Using a Single-View RGB-D Video” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jiahui Wang, Wanqing Du, Yuanbin Zou, Zhefeng Xu, Feng Xu
HOForce estimates fingertip contacts and forces from single-camera RGB-D video, using known object geometry and physical properties to refine hand and object motion. It jointly simulates interacting objects and updates contact and force estimates against observed motion. On 32 recorded sequences, its sequence-balanced normal-force MAE was 1.002 N versus HOPE’s 1.326 N; these are feasible estimates, not uniquely identified forces.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.